Many or most single mode communications-grade optical fibers and many multi-mode fibers are fabricated from high-silica glass components. Such fibers have a high Young's Modulus, and are termed nearly “perfectly elastic” in addition to possessing very low thermal coefficients of expansion. This combination of properties makes the optical fiber quite stable for communications purposes in the field if precautions are taken to protect it from moisture-caused static fatigue failure, hydrogen diffusion (causing higher absorption of light) and physical forces, among other dangers. Such protection means include, but are not limited to, coating (e.g., during the fiber drawing process) with materials such as acrylates, polyimides, carbon, diamond-like carbon, copper, aluminum and other materials that can be applied to the fiber during the high speed drawing process. These coatings are usually termed “buffer” coatings. Subsequently, the fibers are frequently cabled or jacketed with materials that include strength members (e.g., Kevlar fibers) and jackets for crush and kink protection.
Such fibers often include in their structures at least one core with at least one index of refraction and at least one glass cladding adjacent to the core with at least one index of refraction that is lower than that of the core in order to substantially confine light to the core.
Optical fiber sensors of temperature and/or strain based on common fiber Bragg gratings (“FBGs”) can be fabricated in the cores of optical fibers by various means. These gratings are characterized by alternating regions of index of refraction value along a longitudinal length of the fiber core having some pitch, or period. There are several distinct types or varieties of FBGs, including but not limited to short period, long period, blazed and phase shifted gratings. Further, these types can be modified by varying the period (chirp), amplitude (apodizing), index background level and/or physical damage level used to fabricate the gratings. Such damage can be induced by a higher intensity of the FBG fabricating light (usually ultraviolet, or UV lasers; in some cases CO2 or other sources) than is actually necessary to write the grating. The number of cores, core shapes, number of cladding layers, and addition of stress-inducing members can all be varied to control the optical properties for various applications. Different elements can be added to the glass formulation to control the index contrast between the core(s) and the cladding(s).
Advantages of optical sensors over electronic sensors are generally well known, in spite of their present overall greater cost (including the sensor readers). Such advantages include, but are not limited to, immunity to electromagnetic interference (EMI) and electromagnetic pulses (EMP), corrosion resistance, explosion-proof nature, light weight, small size and potential for all-dielectric construction (leading to high voltage compatibility). In addition, sensors based on FBGs enjoy the ability to be multiplexed on a single optical fiber in large numbers by several means, including wavelength division multiplexing (WDM) and optical frequency domain reflectometry (OFDR), leading to a lower cost per sensing point when the cost of the reading instrument is averaged over the number of sensors attached. Further, only a single feedthrough point through bulkheads and pipes is needed for a high sensor count, leading to enhanced ease of installation and lower vulnerability to breach of the bulkhead integrity at the feedthrough. In order to be multiplexed in this way, physically in series along the fiber, the sensors should generally be optically double ended, or have an input fiber and an output fiber (it is understood that the input and output fibers are interchangeable for an FBG). In order to make FBG sensors both small enough to be compatible in form factor with electronic sensors and optically double ended requires innovation beyond the present state of the art.
Most types of FBGs are sensitive to both temperature and strain variables to essentially the same degree for a given type, although the degree of interdependence on the two variables may vary from type to type. Further, if the FBG is fabricated in the core of a high-silica fiber, such as is commonly done, the sensor also has the properties of high Young's Modulus and low coefficient of thermal expansion. These properties generally cause difficulty if the sensor is to be used over a very wide temperature range, if their temperature sensitivities or temperature ranges need to be enhanced beyond that of the simple buffered fiber (by attachment to a material of a higher expansion coefficient), if they will be subjected to rough handling, or firmly mounted to dissimilar materials (to enhance thermal equilibrium with the object to be measured). In addition, fabrication difficulties increase when the effects of strain are to be separated unambiguously from those of temperature and when the sensor is made compact enough to compete with existing electronic sensors in form factor while still maintaining their ability to be multiplexed.
If a section of optical fiber containing an FBG is attached to another object or material (substrate) with adhesive or even thermal grease, the FBG's temperature calibration and even repeatability is significantly and usually adversely affected by all the components of the attachment system, especially over a temperature range of tens or hundreds of degrees Celsius, because of the strain sensitivity of the FBG. If encapsulated in a material such as an epoxy or another material that is not “perfectly elastic” (i.e., a material that is subject to measurable viscous flow), the mechanical stiffness of the fiber causes the fiber to ‘creep’ or move through the viscous material when stressed by changes in temperature or mechanical causes. This occurs even if the length of the attachment or encapsulation greatly exceeds the length of the FBG itself. In addition, the viscous material itself is often not stable under thermal cycling, especially if it is a glass with a low melting point or is a polymer and its glass transition temperature is exceeded. These effects can lead to variations of temperature calibration of many degrees Celsius from cycle to cycle and even to the loss of optical signal through the gradient-induced breakup of the single reflection peak into multiple peaks (termed accidental chirping, in contrast to the intentional variation of the period of a grating during fabrication).
While it can be very difficult to measure strain without temperature effects, measuring temperature without strain affects can be done with varying degrees of success with appropriate packaging in order to remove the FBG from the effects of stress due to handling or attachment to another object. Although such packaging inevitably increases the dimensions, mass and thermal response times of the FBG sensors, such packaging is necessary to make the sensors of general use in industry. On the other hand, it is extremely desirable to make fiber optic temperature sensor packaging as small and thermally fast as possible, and further to emulate the form factors of commonly used electronic temperature sensors to promote the market acceptance of the newer optical technology in the marketplace.
In order to make the sensors in a physically single ended, ‘probe’ configuration such as is easily done with thermocouples and thermistors, with both fibers coming out of the same end of a small tube or other package, the fiber may be bent in at least a 180° ‘hairpin’ curve in a way that avoids losing significant light transmission (a few tenths of a percent per sensor may be permissible in a sensor array of 100 sensors, for example). Conventional communications-grade optical fibers (e.g., Corning SMF-28) begin losing significant amounts of optical transmission when bent in diameters as large as 30 mm.
FIG. 1 shows example experimental data on power loss from single 180° bends in three types of optical fibers. For a single sensor on a fiber, losses of 50-90% might be tolerated, but if several are to be multiplexed on a fiber, losses of less than 1% are desirable. From the point of view that multiplexing can be highly beneficial for lowest systems costs, fewest fibers and feedthrough points, etc., it is evident that low loss can be important. FIG. 1 demonstrates that common communications fiber with a numerical aperture of much below certain numerical apertures may present challenges for use to make a compact sensor, but if the numerical aperture is increased, smaller sensors can be fabricated that are also low loss. Generally speaking, a starting fiber with any numerical aperture may be used when a section of the fiber is drawn and tapered to an air-guided core (e.g., fiber drawn down to only a few micrometers in diameter). In that case, the index contrast between the glass and the air (n=1) is essentially the index of the glass, or 1.4 to 1.5, so the numerical aperture of the air-guided core fulfills the requirement of very high NA no matter what the starting NA was.
The data shown in FIG. 1 does not address concerns with the well-known increased static fatigue failure of optical fiber as the bend radius decreases below about 3 mm; it is an illustrative example only of the optical power loss due to the bend. For example, the estimated lifetime of a buffer-coated fiber (that has never been stripped and recoated) bent in a 3 mm radius is greater than 50 years while a bend of 1.5 mm radius would have a failure time measured in hours. To avoid any such increase in fatigue failure for bend radii less than about 2 to 4 mm, thermal bending and annealing of the fiber comprising the bend is desirable. If the thermal process does not induce an adverse chemical or physical change in the core of the fiber, the benefits of the high numerical aperture fiber will be retained and the probability of fiber fracture will be significantly reduced.
In general use, a fiber optic sensor package with a width or diameter of 20-30 mm or greater is highly undesirable. Since electronic industrial sensors frequently are packaged in tubes with diameters of 0.5 to 13 mm, optically double ended, physically single ended fiber optic temperature sensor probes with diameters of 0.3 to a maximum of 13 mm, and preferably 0.3 to 6 mm, will find enhanced utility in industry. This discussion of round or tubular sensor probes does not exclude other cross sectional geometries, such as rectangular or oval cross sections.
The exemplary illustrative technology herein provides compact, optically double-ended sensor probes with at least one substantially 180° bend provided in the optical fiber in close proximity to an FBG sensor. This example non-limiting structure may include for example all versions of at least net 180° bends in definition and bends of somewhat less than 180° that would lead to slightly divergent input and output fibers but still allow a physically single-ended probe configuration within a desired maximum diameter. Further, the FBG sensor can in example non-limiting implementations be suspended in the probe in such a way that the expansion and contraction of the probe casing will not materially influence the temperature reading of the FBG by adding time-or-temperature varying stress components to the FBG. Such time-dependent drift mechanisms that can be avoided include creep in reading (at a constant temperature) that frequently occurs when attempts are made to fasten fibers incorporating FBGs at both ends of the FBG to the casing in a direction substantially on a line with each other, even if said fiber is bent somewhat (substantially less than 180°) to prevent fiber breakage.
Mechanical 180° bends can be mechanically restrained to force them into a compact form factor if means are employed to prevent such restraints from themselves causing variations in the calibration of the sensors with time and temperature cycling. Thermally formed bends can be made by heating the fiber beyond its softening point utilizing any of the methods of, but not exclusively confined to, a flame, an oven, a hot filament, a glow bar, or a laser, for instance a CO2 laser. The buffer coatings can be removed before heating, burned off during the bending operation or, if an inert atmosphere is employed, an adherent, protective carbon layer can be left on the fiber bend. Reliability of the bend can be enhanced by annealing and slow cooling the bend. Since FBGs in many fibers can be erased by high temperature, the FBG can be of a type that can withstand the temperature of the bending operation, it can be written into the fiber before bending and kept a safe distance away from the bend or the fiber can be loaded with hydrogen (if required) after the bending operation and the grating can be written into the bent fiber after the hydrogen loading step.
Additional exemplary illustrative non-limiting features and advantages include:                A compact optical fiber temperature sensor that is optically double ended and can be made either physically single or double ended as a probe or in-line, respectively, encompassing at least one FBG in close proximity to at least one bend in the fiber comprising at least one net 180° path        said fiber having a numerical aperture of much above a certain numerical aperture such as about 0.11 in one particular non-limiting example        said FBG further mounted within an outer casing        the optical fiber is a single mode fiber        the radius of the largest at least one net bend is from 0.01 mm to 10 mm and preferably from 0.15 mm to 5 mm        an arrangement of fiber and FBG is mounted and maintained in physical independence of expansion and contraction of the outer casing, including rubbing on the casing.        mechanical stress placed upon the outer casing, as in fastening said casing to an object to be measured for temperature, has no or substantially no effect on the temperature calibration of the FBG        the input and output fibers emerge or can be caused to emerge from the casing essentially at the same end and substantially parallel to each other in a probe configuration        the application of a bend or bends of greater than 180° or multiple bends can cause the input and output fibers to emerge from the casing at up to 180° from each other (i.e., at 90° or 180° substantially orthogonal or parallel to each other)        said casing contains an atmosphere        said casing is environmentally and/or hermetically sealed        the at least one FBG is contained in a straight section of fiber within the distance of from 0.01 mm to 100 mm from one end of the at least one 180° bend in the fiber and preferably from 1 mm to 10 mm        the at least one FBG is at least partially contained within the at least one net 180° bend in the fiber        the FBG is fabricated in un-stripped fiber according to the definition        the at least one 180° bend in the fiber is formed by at least one method chosen from the group mechanical bend, thermal bend or tapered bend (per the definitions)        the at least one bend contains at least one length of longitudinally tapered index of refraction, exclusive of the FBG        the at least one net 180° bend in the fiber is confined and supported by at least one rigid band across approximately the diameter of the bend and weighing between 10 ng and 10 g, and preferably between 10 μg and 100 mg, such that the weight of the at least one band is supported entirely by the fiber, free of contact with the casing, and further the at least one band moves freely with the fiber within the sensor casing without adding variable stress to the FBG        the casing is made entirely of dielectric materials        the casing is hermetically sealed        the hermetic seal is chosen from one of the group of a weld, a metal alloy solder or sealing glass composition        the optical fiber in the area of the hermetic seals is provided with at least one of or a combination of a solderable metal coating, an organic buffer coating or no buffer coating        the metal alloy solder contains at least one rare earth element.        the FBG is contained within a 360° bend        the 360° bend is substantially circular and the diameter of said circle of fiber is fixed by at least one support at the point of closure of said circle of fiber and which further forces the input and output fibers to emerge substantially tangentially to and near the plane of said circle        the length of the FBG is bare (without a buffer coating)        the FBG is protected from damage by at least one support with at least one projection running substantially parallel to the FBG and weighing between 10 ng and 10 g, and preferably between 1 μg and 10 mg        said longitudinal support is a tube encompassing the FBG and unattached to the fiber (i.e., floating freely on the fiber)        said tube is attached to the fiber or other portion of the support structure on only one end, and in which the fiber is free to expand and contract independently of said tube        the tube is composed of one or more of a metal, metal alloy, glass, ceramic, composite or polymer        the atmosphere in the casing contains helium gas for the purpose of enhanced thermal conduction between the casing and the FBG        the optical fiber containing the at least one 180° bend is holey fiber, nanostructured fiber or photonic crystal fiber        a bend in an optical fiber in which the optical intensity losses are reduced by increasing the index of refraction of the fiber core within only the bend by means of exposure to ultra violet radiation at least over a portion of the length of the bend        a bend in an optical fiber in which optical fiber within the bend is subjected to ultraviolet radiation in order to increase its numerical aperture, in which process said deep ultra violet radiation is any combination of constant and varying intensity over the length of the bend        
Note: The drawings herein represent the fiber in two dimensions while assuming a buffer coating is included on the fiber except as noted below.